Sensitivity of Radial Endobronchial Ultrasound-Guided Bronchoscopy for Lung Cancer in Patients With Peripheral Pulmonary Lesions: An Updated Meta-analysis

Background: Registry trials have found radial endobronchial ultrasound (r-EBUS) sensitivity to vary between institutions, suggesting that in clinical practice, r-EBUS sensitivity may be lower than reported in clinical trials. We performed a meta-analysis to update the estimates of r-EBUS sensitivity and to explore factors contributing to heterogeneity of results. Methods: A systematic review using PubMed was performed through July 2018 to determine the sensitivity of r-EBUS for lung cancer, and to construct a summary receiver operating characteristic curve. The DerSimonian and Laird method was used to weight results. Subgroup analysis and meta-regression was used to identify sources of heterogeneity. Study quality was assessed using the QUADAS tool, and publication bias was tested using funnel plots. Results: Fifty-one studies with a total of 7,601 patients were included. r-EBUS pooled sensitivity was 0.72 (95% CI, 0.70-0.75), and area under the sROC curve was 0.96 (95% CI, 0.94-0.97). Significant heterogeneity was observed (I² = 76%; heterogeneity P <.01). We failed to demonstrate an association between sensitivity and air bronchus sign, average nodule size, use of fluoroscopy, virtual bronchoscopy, guide sheath, cancer prevalence, multicenter status, or consecutive enrollment. Rapid onsite cytology was associated with increased sensitivity (P =.01). The pooled pneumothorax rate was 0.7% (95% CI, 0.3%-1.1%). Funnel plots were asymmetrical, demonstrating sample size-related effects and possible publication bias. Conclusions: r-EBUS has an excellent safety profile, but there is significant between-study heterogeneity. Sample size-related effects and possibly publication bias have led to overly optimistic estimates of the sensitivity of r-EBUS.